Run verification commands and fix failures until quality goals pass.
npx skills add https://github.com/xdkp/oh-my-claudecode --skill ultraqa-xdkp
Please help me install this Agent Skill. Skill: ultraqa Source: https://github.com/xdkp/oh-my-claudecode/tree/main/skills/ultraqa Command: npx skills add https://github.com/xdkp/oh-my-claudecode --skill ultraqa-xdkp
Ultraqa eliminates the frustration of repeatedly running tests, interpreting failures, and manually hunting down fixes by coordinating an end-to-end QA loop until your quality goal is satisfied.
Use case: You ship a change to a TypeScript repo and need confidence that tests, linting, and type checking all pass; Ultraqa iterates through failing suites, diagnoses the root cause, applies fixes, and repeats until the checks succeed.
Run /oh-my-claudecode:ultraqa with the appropriate flag (for example, --tests) to start a QA cycle and iterate until the verification goal is met.
High-intent search queries and answers about installing and using this skill.
You can automate TypeScript verification cycling by repeatedly executing your project's build, lint, typecheck, or test commands and applying iterative fixes until the defined quality goal succeeds.
Autonomous QA cycling runs verification commands iteratively; on failure, it spawns an architect subagent to identify root causes and an executor subagent to apply precise fixes, tracking state across up to five cycles until checks pass.
To set up an auto-fix workflow, run the QA cycle with the appropriate flag for your verification goal, such as tests or lint. The system will iterate, diagnose failures using subagents, and apply fixes up to a maximum of five cycles.
Yes, automated diagnosis and repair applies to custom output-pattern verification alongside standard tests, builds, lint, and type checking, iterating through failures and applying fixes until your defined quality goal passes.
When automated test fixing detects repeated failures across cycles, it stops early with a focused diagnosis to prevent wasted iteration, as the system tracks failure state and applies guardrails against infinite loops.
Automated QA cycles limit iteration to a maximum of five cycles per run, spawn architect and executor subagents only on failures, and require parsing goal arguments to execute the corresponding verification commands.
nichobbsAutomate iterative QA loops with failure diagnosis and fix application.
learntocloudRun pre-commit checks, stage, commit, push, and monitor deployment.
miko06Compute code quality scores from type checkers, linters, tests, and dead code detectors.
jMertaDiagnose and patch failing GitHub Actions CI workflows using the GitHub CLI.
